Filters








14,474 Hits in 5.4 sec

Predicting Pedestrian Crossing Intention with Feature Fusion and Spatio-Temporal Attention [article]

Dongfang Yang, Haolin Zhang, Ekim Yurtsever, Keith Redmill, Ümit Özgüner
2021 arXiv   pre-print
This work addresses the above limitations by introducing a novel neural network architecture to fuse inherently different spatio-temporal features for pedestrian crossing intention prediction.  ...  Predicting vulnerable road user behavior is an essential prerequisite for deploying Automated Driving Systems (ADS) in the real-world.  ...  ., c t g } offers the visual features that account for multi-interactions between the road and road users, or among road users.  ... 
arXiv:2104.05485v2 fatcat:o5efy5n2cbcdfi7xsbclabegpq

Multi-Classifier Feature Fusion-Based Road Detection for Connected Autonomous Vehicles

Prabu Subramani, Khalid Nazim Abdul Sattar, Rocío Pérez de Prado, Balasubramanian Girirajan, Marcin Wozniak
2021 Applied Sciences  
feature fusion and a deep learning classifier.  ...  In this paper, a two-level fusion of classifiers (TLFC) approach is proposed by using deep learning classifiers to perform accurate road detection (RD).  ...  The output from the cell is used as a feature map for road detection in CAVs.  ... 
doi:10.3390/app11177984 fatcat:fsvnlk4ekzg2bcusw6wx32kkwi

Bayesian Information Criterion Based Feature Filtering for the Fusion of Multiple Features in High-Spatial-Resolution Satellite Scene Classification

Da Lin, Xin Xu, Fangling Pu
2015 Journal of Sensors  
In the second phase, Bayesian information criterion (BIC) is utilized to conduct the feature filtering which sets the smallest loading in absolute value to zero in each iteration for all features.  ...  This paper presents a novel classification method for high-spatial-resolution satellite scene classification introducing Bayesian information criterion (BIC)-based feature filtering process to further  ...  In this paper, the -fold cross validation approach is used to optimize the penalty parameters for each canonical feature pair, where in -fold cross validation is set to 10 by empirical.  ... 
doi:10.1155/2015/142612 fatcat:tfmjlzy43nfalhls5wpykgvudq

Pedestrian Stop and Go Forecasting with Hybrid Feature Fusion [article]

Dongxu Guo, Taylor Mordan, Alexandre Alahi
2022 arXiv   pre-print
Considering the lack of suitable existing datasets for it, we release TRANS, a benchmark for explicitly studying the stop and go behaviors of pedestrians in urban traffic.  ...  We also propose a novel hybrid model that leverages pedestrian-specific and scene features from several modalities, both video sequences and high-level attributes, and gradually fuses them to integrate  ...  for the task of recognizing pedestrians' intentions of crossing the roads [44] .  ... 
arXiv:2203.02489v1 fatcat:auat4itsmfg47oqg6lvdfeu27y

Multi-Feature Fusion and Enhancement Single Shot Detector for Traffic Sign Recognition

Yanmei Jin, Yusheng Fu, Wenqin Wang, Jinhong Guo, Chunhui Ren, Xin Xiang
2020 IEEE Access  
In this paper, we propose an improved (Single Shot Detector) SSD algorithm via multi-feature fusion and enhancement, named MF-SSD, for traffic sign recognition.  ...  INDEX TERMS traffic sign detection; small target detection; single shot detector; feature fusion; feature enhancement This work is licensed under a Creative Commons Attribution 4.0 License.  ...  Figure 3 shows the process of feature fusion from conv4_3 to the feature group to be detected.  ... 
doi:10.1109/access.2020.2975828 fatcat:jtvgf7rn6jd2df2acbdz7at67m

Human detection using multimodal and multidimensional features

Luciano Spinello, Roland Siegwart
2008 2008 IEEE International Conference on Robotics and Automation  
This paper presents a novel human detection method based on a Bayesian fusion approach using laser range data and camera images.  ...  The selection of HOG features and laser features is obtained through a learning process based on a cascade of linear Support Vector Machines (SVM).  ...  2) increasing the feature set of the previous work of Arras [2] for range based human detection, considering 2D features and n-dimensional features (shape factors). 3) the implementation of a novel fast  ... 
doi:10.1109/robot.2008.4543708 dblp:conf/icra/SpinelloS08 fatcat:o4yjhucwtrc2hodahfe6vz6l5i

Deep Fusion of DOM and DSM Features for Benggang Discovery

Shengyu Shen, Jiasheng Chen, Shaoyi Zhang, Dongbing Cheng, Zhigang Wang, Tong Zhang
2021 ISPRS International Journal of Geo-Information  
The results show that the fusion of DOM and DSM data is beneficial for benggang detection via supervised convolutional and deep fusion networks.  ...  The two sources of information (DOM and DSM) are fused via a gated neural network, which learns the most discriminative features for the detection of benggang.  ...  Acknowledgments: The authors would like to thank the reviewers and editors for valuable comments and suggestions. Conflicts of Interest: The authors declare no conflict of interest.  ... 
doi:10.3390/ijgi10080556 fatcat:6n74vgrfwbb5tkrbrj4up25lsy

MASFF: Multiscale Adaptive Spatial Feature Fusion Method for Vehicle Recognition

Jiefei Zhang Jiefei Zhang
2022 Diànnǎo xuékān  
In this paper, we propose a multiscale adptive spatial feature fusion (ASFF) method for vehicle recognition. First, it calculates the difference hash values of images.  ...  Meanwhile, multi-scale adaptive spatial feature fusion method is adopted to fuse the multi-level features of the network.  ...  spatial feature fusion (MASFF) method for vehicle recognition.  ... 
doi:10.53106/199115992022023301001 fatcat:4kojtye2vnc3npa424cqtbvdjy

Improve SegNet with feature pyramid for road scene parsing

Xinbo Ai, Yunhao Xie, Yinan He, Yi Zhou, S. Trukhanov, C.-W. Chow
2021 E3S Web of Conferences  
We extend the segmentation network with an encoder-decoder structure by adding an edge feature pyramid module, namely Edge Feature Pyramid Network (EFPNet, for short).  ...  Road scene parsing is a common task in semantic segmentation.  ...  In the module, after each convolutional layer of SegNet, a 1×1 convolutional layer is used for dimension reduction.  ... 
doi:10.1051/e3sconf/202126003012 fatcat:6xo3gza2fbdlzhjwku5hjiquba

Feature Pyramid Transformer [article]

Dong Zhang, Hanwang Zhang, Jinhui Tang, Meng Wang, Xiansheng Hua, Qianru Sun
2020 arXiv   pre-print
To this end, we propose a fully active feature interaction across both space and scales, called Feature Pyramid Transformer (FPT).  ...  We conduct extensive experiments in both instance-level (i.e., object detection and instance segmentation) and pixel-level segmentation tasks, using various backbones and head networks, and observe consistent  ...  Acknowledgements We would like to thank all the anonymous reviewers for their constructive comments. This  ... 
arXiv:2007.09451v1 fatcat:s5uccxm52bclbjanhbbeeog2ha

Scale-Sensitive Feature Reassembly Network for Pedestrian Detection

Xiaoting Yang, Qiong Liu
2021 Sensors  
In this paper, we propose a novel scale-sensitive feature reassembly network (SSNet) for pedestrian detection in road scenes.  ...  Serious scale variation is a key challenge in pedestrian detection. Most works typically employ a feature pyramid network to detect objects at diverse scales.  ...  Conclusions In this paper, we propose SSNet, a scale-sensitive feature reassembly network, for handling severe scale challenges in pedestrian detection under road scenes.  ... 
doi:10.3390/s21124189 fatcat:o3glrsk5mrb4fl2bs6k53c7fsq

(AF)2-S3Net: Attentive Feature Fusion with Adaptive Feature Selection for Sparse Semantic Segmentation Network [article]

Ran Cheng, Ryan Razani, Ehsan Taghavi, Enxu Li, Bingbing Liu
2021 arXiv   pre-print
We present a novel multi-branch attentive feature fusion module in the encoder and a unique adaptive feature selection module with feature map re-weighting in the decoder.  ...  Recently, several methods have been introduced for 3D LiDAR semantic segmentation.  ...  matrix with K feature dimensions.  ... 
arXiv:2102.04530v1 fatcat:ixefmjclzrabllhqlle4sohlae

Dynamic Feature Fusion for Semantic Edge Detection [article]

Yuan Hu, Yunpeng Chen, Xiang Li, Jiashi Feng
2019 arXiv   pre-print
In this work, we propose a novel dynamic feature fusion strategy that assigns different fusion weights for different input images and locations adaptively.  ...  This is achieved by a proposed weight learner to infer proper fusion weights over multi-level features for each location of the feature map, conditioned on the specific input.  ...  The ground truth maps are downsampled into half size of original dimensions for Cityscapes, and are generated with instancesensitive edges for both datasets.  ... 
arXiv:1902.09104v1 fatcat:aucuf7npvvc33lm5bkbketvehm

Radar-Vision Fusion For Vehicle Detection By Means Of Improved Haar-Like Feature And Adaboost Approach

Anselm Haselhoff, Anton Kummert, Georg Schneider
2007 Zenodo  
For example, an adaptive MI threshold or a cross-validation set could be used.  ...  We tested this approach in rectangle feature selection for vehicle detection.  ... 
doi:10.5281/zenodo.40629 fatcat:rkaqp5sw3jefjgttwf2tokduni

Automated Pavement Crack Damage Detection Using Deep Multiscale Convolutional Features

Weidong Song, Guohui Jia, Hong Zhu, Di Jia, Lin Gao
2020 Journal of Advanced Transportation  
Road pavement cracks automated detection is one of the key factors to evaluate the road distress quality, and it is a difficult issue for the construction of intelligent maintenance systems.  ...  Compared with other state-of-the-art methods, the CrackSeg performs more efficiently, and robustly for automated pavement crack detection.  ...  Supplementary Materials CrackDataset is an annotated road crack image database which can re ect road surface condition in general. (Supplementary Materials)  ... 
doi:10.1155/2020/6412562 fatcat:aeqx3iy44bhilm2ggq67qlxgze
« Previous Showing results 1 — 15 out of 14,474 results